I've seen a few threads now starting to talk about very long timescale MMORPGs, setting the game time period 'from primitive man to spacefaring' and so on. There are a number of reasons why these ideas are not viable for an MMORPG setting (though they can more easily be adapted to an online strategy game). In any case, they're exceedingly large projects. Let's first consider assets: Artwise, RPGs, especially those that allow free-roaming (like most MMOs) are exceedingly asset-heavy. 'Levels' (or zones, whatever) have large numbers of textures associated with them, characters have a wide range of equipment in a wide range of styles all of which require their own assets, artist designed - even using modular construction of items requires artist input at the component level. This is assuming a fixed time period (where 'fashion' doesn't change too much). If you expand this to include an evolutionary expansion of technology, architecture and fashion you end up needing to develop a *huge* amount of assets. Simply consider the changes seen in the last century - the different styles of clothes, of cars, of guns, of architecture... Considering gameplay: What is the role of the player in an evolutionary game? It's necessary to effectively accelerate the flow of time, at least as far as technological advancement is concerned - but it would be strange to see a character that starts as a caveman fly a starship. How should character advancement work - what use is a flint spear skill to a future star pilot? Why should a level 40 caveman be better at flying a starcruiser? What happens when the endgame for a planet (future tech) is reached? Does the game reset? What happens to player advancement? Where is the 'hook' to carry on playing? It seems that a lot of these ideas are falling off the back of civilisation style concepts without realising that the player role in that sort of game is a detached 'god'. Applying the evolutionary concept to a single player's world in a massively multiplayer galaxy / universe (or indeed species as in Spore) is much more viable - providing the disparity in available tech is taken into account in PvP rules, ie. spacefaring civilisations should not attack cavepeople. It should be borne in mind that a detached 'god' role requires FAR less detail on the part of the subject civilisation than a immersed 'you're one of them' role, such as you'd expect in an MMORPG. Whole epochs of history can be flipped between (civ style) without ruining the immersion. Altering the perceived speed of time (as far an immersive game is concerned) should be done with EXTREME caution. It's my opinion that progressive 'end-game' plots should not be used to drive MMO's (or any RPG). Allowing the balance of a gameworld to become disrupted causes economic, political and social disruption, much as happens in the real world. Since these games are meant to be fun, disruption should happen in short bursts, to present a challenge, and should be *controlled* with the intention of providing your player base with a fun experience.

The closest thing i've seen to this concept of civilization advancement in a MMOG is 'A Tale in the Desert'. It's an indie developed game, (I actually helped out with some terrain building back in alpha [smile]) the smartest thing it's developers ever did was identify the limit's of the project. They understood that art assets needed to be simple and re-usable and that they would need a powerful script system to make the constant development of technology (new items, materials etc) feasible.

Even so, the game only spans a period of a few hundred years before a reset - it's up to the third version now I think. This model is strongly episodic, with sizable game updates with each new incarnation, which probably helps to bring back old players for another try.

I think there is certainly room for experimentation with timescales, but as you point out, there are some serious technical and gameplay issues to overcome. From a design perspective I expect this family management idea (which is oft discussed on these boards), with base skills and wealth passed from generation to generation, would be the obvious choice to keep character skills relevant to the age. Other than that, I wouldn't like to think to hard about it, after all, it is a completely mad thing to try to build.

One thing is certain. You can't develop an evolutionary style MMO with static assets and deployment.

Then again, assets don't have to be static. You could have an MMO in which players could start off at varying stages of civilization on different planets/galaxies, and where the development of a civilization would differ based on the players that comprise it. Players could jump in at the baseline, or start in an already grown civilization. The replay value comes in the form of being able to explore other developing civilizations and start new civilizations anew. After that comes conflict, or perhaps agreement and trade.

The only problem is that such an MMO would take forever to design with static assets and would be impossible to maintain - with static assets. Of course, if you did everything procedurally, you wouldn't have those problems. Most of the content (civilizations) would be player created - it'd be kinda like an MMO spore. Essentially, the game is seeded with a few civilizations of various size/developement/technology for new players to jump into. As the players play the game, whatever type of game it may be, they grow their civilization and accelerate through flexible periods of development. The key here is to have generalized instead of specific technologies that tie into your procedural development system. E.g., instead of having to research "space travel" or "rocketships" the players must access things like "particle physics" and try and build ships and technologies based on a wide variety of technological discoveries. The development team builds prototypes, everything else is player created. You build a "particle emitter" and "phosphoric metal alloys" and the player puts them together into a space ship. Spore style. The player discovers "metalite" and "stalactacin," each having specific properties. He heats them and combines them into another substance, "metalactacin," which has properties dependant on the first two. All you do is define the properties of base materials and the procedure for their interaction under certain situations, and the player builds their own world.

The more variation you have in the procedure, the more replayability and the more there is to discover. In a game like this, the limit to "technology" is as high as you set it in terms of base materials and possible interactions. You can define base actions like "sharpen" and "heat" for the player to use, and even these actions can be based on simpler actions. Lots of coding, little content creation, infinite playability.

It seems that most if not all raw geometry in Spore is artist designed. The buildings, the UFO, the interface, the components for the varying creature parts. All have complex rigs to allow deformation and skeletal stitching. Procedural techniques are apparently limited to selection and blending of components, and generation of appropriate animations based on the composite rig. Until I see a better video, or sit down and play it for a few hours I couldn't analyse it much more than this.

Procedurally generating machines and buildings for civilisations still requires a whole set of base components, and construction rules for each machine type, most likely hierarchically defined, eg: wheeled_ground_vehichle->car, wheeled_ground_vehicle->train->locomotive, wheeled_ground_vehicle->train->rolling_stock->passenger carriage.

To simply present the raw components required to (graphically) construct everything required for a game of this ilk (evolutionary) is substantially more work than to construct appropriate components for a 'normal' MMO.

Incidentally, component based entity construction is one of the design tools featured in Primogen (my MMO toolset), for precisely the reason Nytehauq stated - variability and adaptivity.